Research map / Image generation / Text-to-image diffusion
Control Color: Multimodal Diffusion-Based Interactive Image Colorization
Zhexin Liang, Zhaochen Li, Shangchen Zhou, Yu Li, Chen Change Loy
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add deformable decoder (ablation) | CtrlColor unconditional · ImageNet val5k | FID: 9.7989 -> 8.3047 | Improved |
| add deformable decoder (ablation) | CtrlColor unconditional · ImageNet val5k | PSNR: 19.4400 -> 21.5241 | Improved |
| add deformable decoder (ablation) | CtrlColor unconditional · ImageNet val5k | Colorfulness: 59.8463 -> 28.6752 | Hurt |
| add SAG (ablation) | CtrlColor unconditional · ImageNet val5k | FID: 9.7989 -> 11.2395 | Hurt |
| add SAG (ablation) | CtrlColor unconditional · ImageNet val5k | LPIPS: 0.2414 -> 0.2332 | Improved |
| add SAG on top of deformable decoder (ablation) | CtrlColor stroke-based · ImageNet val5k | FID: 5.0103 -> 4.6657 | Improved |
| add SAG and deformable decoder (ablation) | CtrlColor stroke-based · ImageNet val5k | PSNR: 26.6739 -> 28.4350 | Improved |
| add deformable decoder on top of SAG (ablation) | CtrlColor unconditional · ImageNet val5k | FID: 11.2395 -> 8.8749 | Improved |
| use streamlined instead of original SAG (ablation) | CtrlColor · ImageNet val5k | PSNR: 21.0038 -> 21.0409 | No clear effect |
| increase hint cells from 0-20 to 50-100 (ablation) | CtrlColor hint · COCO-Stuff | FID: 8.8614 -> 6.8830 | Improved |
| increase hint cells from 0-20 to 50-100 (ablation) | CtrlColor hint · COCO-Stuff | PSNR: 24.4910 -> 28.2699 | Improved |
| increase hint cells from 0-20 to 50-100 (ablation) | CtrlColor hint · COCO-Stuff | Colorfulness: 36.3481 -> 32.0113 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- BigColor — BigColor: Colorization Using a Generative Color Prior for Natural Images (ECCV 2022) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Color Alignment in Diffusion (CVPR 2025 · similarity 0.89)
- Automatic Controllable Colorization via Imagination (CVPR 2024 · similarity 0.89)
- Free-Lunch Color-Texture Disentanglement for Stylized Image Generation (NeurIPS 2025 · similarity 0.88)
- Color Conditional Generation with Sliced Wasserstein Guidance (NeurIPS 2025 · similarity 0.87)
- Enabling Region-Specific Control via Lassos in Point-Based Colorization (AAAI 2025 · similarity 0.87)
- Image Referenced Sketch Colorization Based on Animation Creation Workflow (CVPR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
